Research article Spatial and Temporal Trends of PM₁₀ and SO₂ in the Area

Nick O. Okello*1, Sandy Camminga2, Tom W. Okello1 and Mark Zunckel3 1University of Free State, Private Bag X13, Kestell Rd, Phuthaditjhaba, 9866, Free State, , [email protected] and [email protected] 2Richards Bay Clean Air Association (RBCAA), Richards Bay, KwaZulu-Natal, South Africa, [email protected] 3uMoya-NILU Consulting, Durban, KwaZulu-Natal, South Africa, [email protected]

Received: 26 October 2018 - Reviewed: 31 October 2018 - Accepted: 11 November 2018 http://dx.doi.org/10.17159/2410-972X/2018/v28n2a20

Abstract Air pollution is a public health emergency (WHO, 2016a). It is the biggest environmental risk to health with a global responsibility for about 1 in 9 deaths annually. In 2013, Statistics SA indicated that 10 % of all deaths in South Africa were attributed to respiratory diseases. Areas with increased industrial growth such as Richards Bay are particularly vulnerable. The paper analyses the spatial and temporal concentration trends of PM₁₀ and SO₂ in Richards Bay over the last two decades; specifically, since the promulgation of national ambient air quality standards (NAAQS) and minimum emissions standards (MES). Mann-Kendell trend tests was applied to monitoring data from the Richards Bay Clean Air Association (RBCAA) to test for an improving or worsening trend and the significance thereof. The pollution concentration data was also compared to NAAQS and the WHO annual guidelines. Overall, results indicate that although there has been a downward trend in the concentration of PM₁₀ and SO₂ emissions in the study area, the trends are not statistically significant. However, there was indication of statistically significant reducing trends in ambient concentrations at some stations. The concentrations at all stations were below NAAQS.

Keywords air quality, trends, status, Richards Bay

Introduction Air pollution is a public health emergency with the biggest effects. Pollution trends in growing areas such as Richards Bay environmental risk and carries a global responsibility for about need to be monitored to provide a basis for advice and planning 1 in 9 deaths annually (WHO, 2016a). In 2015, close to 10 % of all for future development and most importantly to protect people deaths in South Africa were attributed to respiratory diseases and the environment. (Stats SA, 2016a). To understand the potential for burden of disease due to air It is known that industrial growth often results in increased pollution, and to ascertain whether policy intervention or emissions and with this a concern for people’s health (Cohen et other drivers have led to an improvement in in air quality, it is al., 2004; Scorgie, 2012). Research has shown that air pollution important to understand the spatial and temporal trends of air exacerbates and increases the probability of infections such pollution. This paper focuses on analyses of two pollutants of as acute lower respiratory infections (ALRI) which include concern PM₁₀, as well as SO₂, and their concentration over the acute bronchitis and bronchiolitis, influenza and pneumonia. last two decades and specifically since the promulgation of In addition, air pollution is associated with stroke, chronic national minimum emissions standards. obstructive pulmonary disease (COPD) including chronic bronchitis and emphysema as well as ischemic heart disease Study area (IHD) and cancers of the trachea, bronchus and lung (Chiu, 2013; Richards Bay is a growing metropolitan with a number of World Health Organization, 2016). industries in relatively close proximity to residential areas (Okello and Allan, 2015). The city is an economic centre and one Whereas accelerated industrial growth is an important aspect to of the country’s strategic economic hubs designated as an ‘Area unlock poverty and improve people’s livelihoods, environmental of National Economic Significance’ (uMhlathuze Municipality, considerations are important to moderate potential adverse 2016). The city hosts the largest coal export terminal in the

CLEAN AIR JOURNAL 80 Volume 28, No 2, 2018 Research article: Spatial and Temporal Trends of PM₁₀ and SO₂ in the Richards Bay Area Page 2 of 10

world, which is the second largest port in South Africa as well irritation of the eyes. Consequently, the inflammation of the as the largest aluminium and iron smelters in Africa. In addition, respiratory tract causes coughing, mucus secretion, aggravation the City also hosts several commercial, light and heavy industrial of asthma and chronic bronchitis and makes people more prone activities such as paper, fertilizer and sugar production, which to infections of the respiratory tract (World Health Organization, collectively contribute to most of the air quality concerns in the 2016). From an ecological perspective, when SO2 combines with region. Sugarcane and forestry burning, pesticide usage and water, it forms sulphuric acid; this is the main component of dust associated with agricultural processes are also common. acid rain.

The City has a population of 364, 473 (Stats SA, 2016b) and an In 2007, an estimated 3.7 % of national cardiopulmonary area of 1233.3 km² (Cogta, 2017). The geographical areas covered disease related deaths and 5.1 % of trachea, bronchus and lung include suburbs such as Esikhaleni, Arboretum, Felixton, cancers in adults older than 30 years were attributed to ambient , Veld n Vlei, as well as the Central Business District air pollution (CSIR, 2015). In 2013, Statistics SA indicated that 10 (CBD) St. Lucia, Brackenham and Harbour West. These areas are % of all deaths were attributed to respiratory diseases. representative of the entire Richards Bay and surroundings. Legislation and policy instruments for air emissions reduction Air quality review From a policy perspective, WHO global emission standards are useful to gauge pollution (WHO, 2005). South Africa too has an Air pollution definition and its effects enabling legal framework that sets out the minimum emission The National Environmental Management Air Quality Act (NEM: standards for point source pollution as well as ambient air AQA, 39 of 2004) defines “air pollution” as any change in the quality standards (DEA, 2010). The Atmospheric Pollution composition of the air caused by smoke, soot, dust (including Prevention Act (APPA) was enacted in 1965 and was repealed fly ash), cinders, solid particles of any kind, gases, fumes, on 1st April 2010 with the National Environmental Management: aerosols and odorous substances. This definition is not far from Air Quality Act (NEM: AQA) which was partially enacted on 9th the WHO, which defines air pollution as a contamination of the September 2005 and fully enacted on 1 April 2010. In order to indoor or outdoor environment by any chemical, physical or protect human health and the integrity of the environment, biological agent that modifies the natural characteristics of the the national ambient air quality standards covering priority atmosphere (World Health Organization, 2016). Consequently, pollutants including PM₁₀ and SO₂ were introduced on 24th air pollution increases the risk of respiratory and heart disease December 2009. Similarly, industrial emission standards were in the population. Both short and long-term exposure to air developed and formally introduced on 22nd November 2012 pollutants has been associated to health impacts. More severe (DEA, 2010). These were revised in 2013 (DEA, 2012). impacts affect the already ill, children, the elderly and poor people who are more susceptible (Chiu, 2013); (Craig et al., Table 1: Ambient PM₁₀ average concentration (µg/m3) NAAQS standard 2007); (WHO, 2013, 2016b). This is also the case in South Africa and WHO guideline adopted from (RBCAA, 2018) and in Richards Bay (Jaggernath, 2013). Source Daily Average Annual average Pollutants of major public health concern include particulate Initial NAAQS[a,b] 120 µg/m3[d] 50 µg/m3[e] matter (PM), carbon monoxide (CO), ozone (O3), nitrogen Current NAAQS [a,c] 75 µg/m3[d] 40 µg/m3[e] dioxide (NO2) and sulphur dioxide (SO₂). While risk is depended on the exposure dose and time, the South African Department WHO[f] 50 µg/m 20 µg/m of Environmental Affairs has declared these pollutants as Notes: criteria pollutants (DEA, 2011). The health risks associated with [a] SA Government Gazette 32816 (24th December 2009) particulate matter of less than 10 and 2.5 microns in diameter in terms of the National Environmental Management: Air (PM₁₀ and PM respectively) is well documented (Cohen et al., 2.5 Quality Act 39 of 2004, [b] Effective from 2010 to 2014 2004; Craig et al., 2008). PM is capable of penetrating deep 2.5 [c] Effective from 2015 into lung passageways and entering the bloodstream causing [d] Not to be exceeded more than 4 times in one year. cardiovascular, cerebrovascular and respiratory impacts (Rai, [e] Not to be exceeded. 2015). There is good evidence of the effects of short-term [f] World Health Organisation guideline exposure to PM10 on respiratory health particularly irritation and acute respiratory disorders, but for mortality, and especially as a Until now, the influence of these policy instruments has not consequence of long-term exposure, PM is a stronger indicator 2.5 been adequately reviewed in the context of a reduction of the (WHO, 2016c). health risks that are attributable to air pollution in Richards Bay. The key question of whether these policy instruments have been Additionally, PM from aerospora is linked to the onset of adequate, and their implementation is producing the desired influenza and allergic cases especially during summer and effect is an important aspect that needs to be investigated. spring (Renaut and Bamford, 2004). SO2 is known to affect the respiratory system and the functions of the lungs and causes

CLEAN AIR JOURNAL 81 Volume 28, No 2, 2018 Research article: Spatial and Temporal Trends of PM₁₀ and SO₂ in the Richards Bay Area Page 3 of 10

Table 2: Ambient SO₂ average concentration (µg/m3) NAAQS standard a result of fossil fuel burning in the residential areas as well as and WHO guideline adopted from (RBCAA, 2018) from industrial and power generation sources (Scorgie, 2012). The State of Air Report for South Africa indicates that PM is still 10 min 1 hour Daily Annual Source Average Average Average average the greatest national cause for concern in terms of air quality due to numerous pollution sources but climatic conditions are NAAQS[a] 500 µg/m3[b] 350 µg/m3[c] 125 µg/m3[d] 50 µg/m3[e] also a major factor (Khumalo, 2016). 191 ppb[b] 134 ppb[c] 48 ppb[d] 19 ppb[e] WHO[e] 20 µg/m³ 20 µg/m³ Areas considered as priority include Highveld Priority Area, Vaal Triangle Priority Air-Shed, and Waterberg Bojanala Priority Notes: Area (DEA, 2018). These areas are highly polluted mainly due to [a] SA Government Gazette 32816 (published 24th December anthropogenic influence. They are declared as priority areas 2009) in terms of the National Environmental Management: due to heavy industrial pollution that often exceeds the NAAQS Air Quality Act 39 of 2004 effective 2010 (DEA, 2018). [b] Not to be exceeded more than 526 times in one year. [c] Not to be exceeded more than 88 times in one year. Despite the implementation of air pollution interventions, [d] Not to be exceeded more than 4 times a year the air quality is said to remain poor and noncompliant with [e] Not to be exceeded ambient standards. One could argue that the Richards Bay area [e] World Health Organisation guideline may have similar characteristics to these priority areas given the number and scale of industrial operations, pollutant sources Table 1 and Table 2 represent the National Ambient Air Quality and history of emission concentration (Renaut and Bamford, Standards (NAAQS) for PM₁₀ and SO₂ respectively as the focus 2004; Scorgie, 2012). Indeed, a study conducted in 2011-2013, pollutants in this paper. focusing on perceptions of air pollution in different suburbs based on economic class concluded that air quality is a priority In this study, only the annual average limit of the national for the residents of Richards Bay in general and poorer areas ambient standards was considered consistent with other are mostly affected by probably as a result of higher solid fuel national and international air quality trend studies and long- burning compared the fuel sources in the more affluent sections term exposure (Mohan and Kandya, 2007; Colette et al., of the area (Jaggernath, 2013). 2011; Guerreiro et al., 2014; Butt et al., 2017). Data is however aggregated from the daily averages and thus diurnal fluctuations have been considered. Methodology Global and local trends in air emission Data collection concentration The ambient air quality data used in these analyses was obtained On a global scale, a study conducted by Butt et al.(2017), from the Richards Bay Clean Air Association (RBCAA). Since 1997, estimates that global population-weighted particulates the RBCAA has grown and established a real time monitoring concentrations increased by 37.5 % over the period 1960 to network able to characterise the local ambient air quality by 2009. Increases in China and India were dominant, attributed to monitoring ground level pollutant concentrations (Okello and economic expansion, and growth in emissions. The study also Allan, 2015). The monitoring network, operated, maintained concluded that in contrast, air quality regulation and emission and managed by an external service provider, comprises 11 controls in the European Union (EU), and United States (US) continuous monitoring stations, recording meteorological data, have reduced population-weighted PM concentrations over 2.5 SO₂, PM₁₀ and Total Reduced Sulphide (TRS) concentrations. the same period.

WHO compared 795 cities in 67 countries for levels of small and Table 3 show the RBCAA stations and surrounding land use fine particulate matter (PM₁₀ and PM ) during the five-year 2.5 areas while Figure 1 shows the spatial locations of the stations. period, 2008-2013. In the study, global urban air pollution levels increased by 8 %, despite improvements in some regions. The monitoring and data collection network is robust and well maintained, with a system in place that satisfies the South Africa Despite very sparse pollution data in the Africa region, the study National Accreditation System (SANAS) requirements (RBCAA, revealed that PM levels remained above the world median 2018). Furthermore, the appointment of an external service (WHO, 2016a). provider allows the RBCAA to remain objective and independent from the results (Okello & Allan, 2015). The RBCAA submit data In South Africa, the air quality in most urban centres is considered to the Department of Environmental Affairs’ South African Air relatively good. However, there are a number of “hotspots” Quality Information System (SAAQIS). This data is also submitted around the country where severe air quality problems are to the WHO Global Ambient Air Quality Database from which experienced (Gwaze, 2017). High ambient SO₂, NO₂ and PM₁₀ WHO derives the global ambient air quality concentrations. concentrations are common in these areas and are primarily

CLEAN AIR JOURNAL 82 Volume 28, No 2, 2018 Research article: Spatial and Temporal Trends of PM₁₀ and SO₂ in the Richards Bay Area Page 4 of 10

Table 3: RBCAA stations and surround area land use Data availability per station was considered. Data used was Predominant filtered to include capture/ availability of over 80% as much as Land use possible where enough data was available. Codes Station Latitude Longitude surrounding station Trend analyses Stn1 Arboretum -28.752704 32.066172 Residential Data was analysed using Mann Kendall test (MK-tests) which has been used widely to statistically analyse for trends in climatologic Stn2 Brakenham -28.731269 32.039014 Residential and in hydrologic time series (Guerreiro, Foltescu and deLeeuw, Stn3 CBD -28.744667 32.055011 Industrial, 2014; Koudahe et al., 2017). This test is used to analyse data commercial and collected over time for consistently upward or downward trends partly residential (“monotonic trends”). The advantage of the MK-tests is that it is Stn4 Esikhaleni -28.865253 31.911651 Residential a non-parametric test. This means it works for all distributions Stn5 Felixton -28.831189 31.895903 Mostly including for data that is not normally distributed. Also the test agricultural, can be used to find trends for as few as four samples (Guerreiro, partly industrial Foltescu and deLeeuw, 2014). Nonetheless, the more data Stn6 Habour -28.787292 32.027125 Industrial points analysed the more the likelihood to find a true trend (as West opposed to chance). The minimum number of recommended Stn7 Mtunzini -28.956439 31.750979 Mostly measurements is therefore at least 8 to 10 (Colette et al., 2011; residential, Karmeshu, 2015; Pohlert, 2018). partly agricultural/ According to this test, the null hypothesis H assumes that there forestry 0 is no trend (the data is independent and randomly ordered) Stn8 Scorpio -28.769697 32.034283 Industrial and this is tested against the alternative hypothesis H1, which Stn9 St Lucia -28.377076 32.414732 Mostly assumes that there is a trend. Data values are evaluated as residential, an ordered time series. Each data value is compared with all partly subsequent data values considered. If a data value from a agricultural/ later period is higher than a data value from an earlier period, forestry the statistic S is incremented by one. On the other hand, if the data value from a later period is lower than a data value sampled earlier, S is decremented by 1. The net result of all such increments and decrements yields the final value of S (Shahid, 2011). The MK-test statistic S is calculated using the formula below (Gilbert, 1987).

(1)

where xj and xk are the annual values in years j and k, j > k, respectively.

(2)

The S statistic is calculated by summing over all pairs of values:

Figure 1: RBCAA Monitoring Network (Source: RBCAA) S will be negative for a negative trend, zero for no trend, and positive for an upward trend. Daily average data per station from 1997 to 2017 was aggregated into annual averages for the nine stations. Seven of these stations For n≤10, the p value is taken from a table of exact values are in Richards Bay within the City of uMhlathuze municipality (Gilbert, 1987). For n>10, a normal approximation is used, while two are in neighbouring municipalities (St. Lucia station in that is by determining the total number of groups of ties g and Mtubatuba and Mtunzini in uMlalazi Municipalities respectively). the number of tied values tj within each group, in the sorted The St. Lucia and Mtunzini stations were considered as reference sequence. Then the standard deviation of S is estimated using sites for PM₁₀ given the predominant North Easterly and South the formula: Westerly wind directions in the area. Seven of the nine stations had PM₁₀ data while six had SO₂ data. (3)

CLEAN AIR JOURNAL 83 Volume 28, No 2, 2018 Research article: Spatial and Temporal Trends of PM₁₀ and SO₂ in the Richards Bay Area Page 5 of 10

The z statistic is then: Particulate matter (PM10) The data analysed indicates that there was a statistically (4) downward trend in ambient PM₁₀ concentrations over last 14 years. Caution was however taken due to the lower number The Z value is then used to calculate p from the cumulative of stations used for sampling prior to 2009. When the same normal distribution. If the p value is less than the significance data was taken from 2009 to 2017, a downward trend was level (alpha) = 0.05, H is rejected. Rejecting H indicates that 0 0 observed but it was not statistically significant. In Table 4, there is a trend in the time series, while accepting H indicates 0 the results indicate that null hypothesis H was accepted for α 0 no trend could be detected. On rejecting the null hypothesis, the 5 out of 6 stations where PM₁₀ data was available. This means result is said to be statistically significant. (Helsel and Hirsch, no significant trends. A downward trend was observed in four 2002; Mohan and Kandya, 2007). stations (Brakenham, CBD, Esikhaleni and Felixton) (Table 4). Only the decrease at Esikhaleni was determined to be statistically significant although sampling had only been a short period of Results and findings time (2014 to 2017). Mtunzini and St. Lucia, the reference site, Overall since 1997, for the Richards Bay area, the PM₁₀ and SO₂ had upward trends but was not statistically significant. The CBD trends examined for significance indicated a reducing trend at and Brakenham had higher PM₁₀ values compared to the other 95 % significance level. However, the trends were not deemed stations, although all measurements were within the stipulated statistically significant using the Mann-Kendall test. Figure 2 NAAQS annual average limit of 50 µg/m³. shows a spatial distribution of the stations and relative PM₁₀ and SO₂ concentrations against the annual average of the NAAQS.

Figure 2: Spatial distribution PM₁₀ & SO₂ annual concentrations in relation to NAAQIS 50µg/m3 annual average.

CLEAN AIR JOURNAL 84 Volume 28, No 2, 2018 Research article: Spatial and Temporal Trends of PM₁₀ and SO₂ in the Richards Bay Area Page 6 of 10

Table 4: Trends of annual mean concentration of PM₁₀ (µg/m³) per station in Richards Bay and in reference locations (2004-2017)

Note: There were no PM10 measurements in Arboretum, Harbour West and Scorpio during the measurement period.

Figure 3: Richards Bay PM₁₀ status and trends (2004-2017)

CLEAN AIR JOURNAL 85 Volume 28, No 2, 2018 Research article: Spatial and Temporal Trends of PM₁₀ and SO₂ in the Richards Bay Area Page 7 of 10

Table 5: Trends of annual mean concentration of SO₂ (µg/m³) per station in Richards Bay and in reference locations (1997 - 2017)

Note: There were no SO2 measurements in St. Lucia and Mtunzini during the measurement period.

Figure 4: Richards Bay SO₂ status and trends (1997-2017)

CLEAN AIR JOURNAL 86 Volume 28, No 2, 2018 Research article: Spatial and Temporal Trends of PM₁₀ and SO₂ in the Richards Bay Area Page 8 of 10

Most importantly, the annual NAAQS was not exceeded in any of were issued. It could also be a result of the implementation of the years where measurements were taken. Although considering minimum emissions standards for listed activities introduced the area average, the CBD, Brakenham and Esikhaleni were in 2010 with achievement milestones stipulated for 2015 and consistently above the average with data predominantly within 2020. The King Cetshwayo District Municipality Air Quality the 90th percentile (Figure 3). Esikhaleni is a highly populated Management Unit and the local uMhlathuze Municipality have area with mostly low-income households and fewer industries incorporated these standards into the Air Emission Licenses compared to areas around the CBD. In this area PM₁₀was likely (AEL) and trade permits of industries. It can be argued that these a result of solid fuel burning. St. Lucia and Mtunzini were the regulatory changes have pushed emission sources particularly reference sites with PM₁₀ levels averaging at 20.8 µg/m³ and industry to invest more on abatement equipment. 22.3 µg/m³ respectively (Figure 3). This is probably a good Another possible factor contributing to the reduction in emission indication of the background PM10 concentration of the whole study area as both sites were relatively unaffected by the main concentration could be the closure of some big industries particularly, Bayside Aluminium, Tata Steel and recently Lafarge sources of PM10. The background in both cases was above the WHO guideline indicating the potential contribution of other that were near the Central Business District. It is worth noting sources such as pollen and sea salts due to proximity to the sea that increase in SO2 and PM10 concentration levels around 2008, (Renaut and Bamford, 2004). 2009 coincide with introduction of industries for example Tata Steel. Also, the Bayside Aluminium and Tata steel closures

An upward trend of PM₁₀ is observed at Mtunzini although not coincide with the sharp decrease in SO2 post 2014. statistically significant. This could be attributed to PM transport due to wind from pollution sources but could also be because of More so, an important contributor to the reducing trend could sea salts, forest fires and vehicles emissions etc. be pressure on industrial emission sources from NGOs such as the RBCAA that has handled numerous complaints from the public and tried to apportion emission to different sources Sulphur Dioxide (SO2) Like the PM₁₀ measurements, SO₂ concentrations in all seven through their stations and case studies. The RBCAA operates stations where data was available were within the stipulated on a polluter pays principle with industry involvement who annual limit of 50 µg/m³ (Table 5). Statistically significant form part of the management committee (Okello and Allan, downward trends were observed in Arboretum, Brakenham, 2015). The RBCAA reviews Environmental Impact Assessments CBD and Felixton. Harbour west neither had an upward nor for new developments in the area that are likely to alter a downward trend. Esikhaleni showed an upward trend emission concentration in the air-shed and advises on potential although with ambient concentrations well below the annual mitigation options. In the words of Jim Phelps, formerly of the limit value. Scorpio had the least favourable SO₂ trends Zululand Environmental Alliance (ZEAL) “the RBCAA has done attributable to their close vicinity to industry while Arboretum, its best to work with polluting industries while raising serious Brakenham, Esikhaleni and Felixton looked better in terms of concerns about air pollution in Richards Bay for the sake of the the SO₂ profile (Table 5). community and environmental health”. Its role thus cannot be under estimated. Data taken over the long term (1997 to 2017) for SO₂ indicate an upward trend although not statistically significant. From The downward trend in ambient concentrations is also mirrored 2013 to 2017 however, a statistically significant downward in the number of air quality related complaints received through trend is observed (Figure 4). The Scorpio and Harbour West the RBCAA and the City of uMhlathuze’s air quality unit which Stations have consistently been above the 20-year average peaked in 2004 and have since showed a downward trend with with annual average ambient concentration within then the 2017 having the least number of complaints (RBCAA, 2018). 90th percentile. This can be attributed mostly to emissions from the surrounding industries. The CBD had SO₂ annual Caution should be taken however, as Richards Bay is rapidly average ambient concentration just below the 20-year regional urbanising with new industrial development underway through annual average. Measurement from residential areas such as the Richards Bay Industrial Development Zones (RBIDZ). The Arboretum, Mtunzini and Esikhaleni showed low concentrations population is also rapidly increasing and with that, more of SO₂. Importantly, the annual NAAQS for SO₂ of 50 µg/m³ was vehicular and other activities that are likely to add to the not exceeded in all stations. PM₁₀ and SO₂ concentration in the ambient air. Monitoring efforts should thus be continued and implemented in a more collaborative and efficient manner to ensure that emission Practical managerial implications sources are reduced for the benefit of all living in the area. and recommendations The results obtained indicate a generally downward trend of ambient PM₁₀ and SO₂ concentrations in Richards Bay over the Conclusion last two decades and particularly in the recent 5 years. This Based on the results analysed, the ambient trends for PM₁₀ and trend could be a result of the introduction of NAAQS in 2009 as SO₂ over the last two decades in Richards Bay and surrounding is seen with the step changes in trends when these standards areas are generally decreasing. This is a positive aspect for

CLEAN AIR JOURNAL 87 Volume 28, No 2, 2018 Research article: Spatial and Temporal Trends of PM₁₀ and SO₂ in the Richards Bay Area Page 9 of 10

the region and could be an indication that efforts to reduce providing data without which this study would not have been emissions by imposing of the national minimum emission possible. standards especially on surrounding industries is achieving the desired effect. However, the trends were not deemed statistically This study is part of a Doctor of Philosophy dissertation through significant using the Mann-Kendall test. Although the overall the University of the Free State, acknowledgment is thus given decrease is not statistically significant, significant downward to the staff and past graduate students for their critique of the SO₂ trends were observed in Arboretum, Brakenham, and CBD. research. For PM₁₀, downward trends were observed in Brakenham, CBD and Felixton with Esikhaleni having the only statistically significant downward trend. References Butt, E. W. et al. (2017) ‘Global and regional trends in particulate This paper thus concludes that efforts to reduce air emissions, air quality and attributable health burden over the past 50 particularly particulate matter in Richards Bay and surrounding years’, Geophysical Research Abstracts, 19(104017), pp. 2017– areas are bearing fruit. Industrial abatement and collaborative 3519. Available at: http://meetingorganizer.copernicus.org/ work through involvement of different stakeholders in forums EGU2017/EGU2017-3519.pdf. such as the RBCAA needs to continue. Government intervention through air emission licensing and trade permits and complaints Chiu, H.-F. (2013) ‘Short-Term Effects of Fine Particulate Air handling is also critical to ensure that pollution reduction trends Pollution on Ischemic Heart Disease Hospitalizations in Taipei: A are sustained to a significant level. Case-Crossover Study’, Aerosol and Air Quality Research, 10, pp. 1563–1569. doi: 10.4209/aaqr.2013.01.0013. It is worth noting that the measured background PM10 in the reference sites of St. Lucia and Mtunzini (20.8 µg/m³ and 22.3 Cogta (2017) ‘KZN Administrative Boundaries’. Cooperative µg/m³ respectively) are in the range of the WHO guideline. These Governance and Traditional Affairs, p. 31. sites were generally not affected by industrial sources and thus could be an indication of other pollution sources mainly sea Cohen, A. J. et al. (2004) ‘Urban Air Pollution’, Comparative salts and probably sugar cane burning. Quantification of Health Risks, 77, pp. 1353–1434. doi: 10.1021/ es304599g.

Recommendation Colette, A. et al. (2011) ‘Air quality trends in Europe over the past It is recommended that the current monitoring of pollution decade : a first multi-model assessment’. doi: 10.5194/acpd-11- concentrations continue to further improve the database of air 19029-2011. quality in the area. Additionally, current efforts by the RBCAA to monitor Total Reduced Sulphur compounds (TRS) should be Craig, L. et al. (2007) Strategies for clean air and health, Journal continued and supported. Measurements should also include of Toxicology and Environmental Health - Part A: Current Issues. other priority pollutants such as nitrogen dioxide, ozone, doi: 10.1080/15287390600882994. benzene, and carbon monoxide where necessary to ascertain baseline and need for continuous monitoring. These additional Craig, L. et al. (2008) ‘Air Pollution and Public Health: A measurements would present a better picture of air quality Guidance Document for Risk Managers’, Journal of Toxicology pollution in the area. Incoming industries with these priority and Environmental Health, Part A, 71(9–10), pp. 588–698. doi: pollutants as emissions should be encouraged to be in involved in 10.1080/15287390801997732. pollution monitoring. More so, recent government involvement in ambient air monitoring should be encouraged and perhaps CSIR (2015) ‘Best Practices Guide. to consider Air-related Human even better consider synergy with existing monitoring networks Health in your Area’. Available at: http://www.saaqis.org.za/. such as the RBCAA to ensure better utilisation of resources and consistent monitoring. DEA (2010) National Environmental Management: Air Quality Act (Act No. 39 of 2004) List of activities which result in atmospheric Although the focus of this paper was on the trends of air quality emissions which have or may have significant detrimental effects pollutants, these results are important and could be applied in on the environment, health, social and ecological conditions. studying health impacts associated with long term air quality South Africa. Available at: https://www.environment.gov.za/. pollution and in interpreting health risk data. The results are also an important indicator of policy efficacy and a reminder DEA (2011) ‘Environmental: Air Quality’, Environmental Outlook, that pollution reduction is possible if policy implantation is pp. 182–207. Available at: http://www.epa.ie/air/quality/. effective. DEA (2012) ‘List of activities which result in atmospheric emissions which have or may have a significant detrimental Acknowledgments effect on the environment, including health, social conditions, Acknowledgment is hereby given to the Richards Bay Clean Air economic conditions, ecological conditions or cultural heritage.’, Association, for the tireless efforts on air quality monitoring in Government Gazette, GN37054(NN964), pp. 3–63. the Richards Bay and surrounding areas and particularly for

CLEAN AIR JOURNAL 88 Volume 28, No 2, 2018 Research article: Spatial and Temporal Trends of PM₁₀ and SO₂ in the Richards Bay Area Page 10 of 10

DEA (2018) South African Air Quality Information Systems Renaut, R. and Bamford, M. (2004) ‘Airspora of the Richards Bay (SAAQIS): Air Quality Priority Areas. Area 2001-2004’, pp. 2001–2004.

Gilbert, R. O. (1987) Statistical methods for environmental Scorgie, Y. (2012) Urban Air Quality Management and Planning pollution monitoring. New York: John Wiley and Sons. Available in South Africa. Doctoral thesis. Department of Geography, at: https://www.osti.gov/servlets/purl/7037501. Environmental Management and Energy Studies, Faculty of Science, University of JohannesburgUniversity of Johannesburg. Guerreiro, C., Foltescu, V. and deLeeuw, F. (2014) ‘Air quality September 2012. status and trends in Europe’, Atmospheric Environment. Elsevier Ltd, 98, pp. 376–384. doi: 10.1016/j.atmosenv.2014.09.017. Shahid, S. (2011) ‘Trends in extreme rainfall events of Bangladesh’, Theoretical and Applied Climatology, 104(3–4), pp. Gwaze, P. (Department of E. A. (2017) ‘South Africa State of Air 489–499. doi: 10.1007/s00704-010-0363-y. Report 2005 to 2016: Presentation’, in. Stats SA (2016a) Community survey 2016. doi: 10.1017/ Helsel, D. R. and Hirsch, R. M. (2002) ‘Statistical Methods CBO9781107415324.004. in Water Resources’, Statistical methods in water resources techniques of water resources investigations, 36(3), p. 524. doi: Stats SA (2016b) ‘Mortality and causes of death in South Africa, 10.2307/1269385. 2015: Findings from death notification’, Statistical release P0309.3, (November), pp. 1–127. doi: Statistical release P0309.3. Jaggernath, J. (2013) ‘A socio-economic and spatial investigation into the health implications of air pollution in Richards Bay, uMhlathuze Municipality (2016) ‘uMhlathuze Municipality KwaZulu-Natal, South Africa’, (March 2013). Annual Report 2015-2016’. Available at: http://www.umhlathuze. gov.za/docs-umhlathuze/annual-report/uMhlathuze-Annual- Karmeshu, N. (University of P. (2015) ‘Trend Detection in Annual Report-Final.pdf. Temperature & Precipitation using the Mann Kendall Test – A Case Study to Assess Climate Change on Select States in the WHO (2005) ‘Air quality guidelines. Global update 2005. Northeastern United States’, Mausam, 66(1), pp. 1–6. Available Particulate matter, ozone, nitrogen dioxide and sulfur dioxide’, at: http://repository.upenn.edu/mes_capstones/47. Environmental Science and Pollution Research, 3(91), pp. 23–23. doi: 10.1007/BF02986808. Khumalo, T. N. (2016) ‘2016 State of the Air Report and National Air Quality Indicator’, (October), p. 45. Available at: https://drive. WHO (2013) ‘Review of Evidence on Health Aspects of google.com/drive/my-drive. Air Pollution—REVIHAAP Project; First Results’, p. 29p. Available at: http://www.euro.who.int/__data/assets/pdf_ Koudahe, K. et al. (2017) ‘Trend Analysis in Rainfall, Reference file/0020/182432/e96762-final.pdf?ua=1%5Cnhttp://www.euro. Evapotranspiration and Aridity Index in Southern Senegal: who.int/en/health-topics/environment-and-health/air-quality/ Adaptation to the Vulnerability of Rainfed Rice Cultivation to publications/2013/review-of-evidence-on-health-aspects-of- Climate Change’, Atmospheric and Climate Sciences, 07(04), pp. air-pollution-revihaapinterim-report. 476–495. doi: 10.4236/acs.2017.74035. WHO (2016a) Air pollution levels rising in many of the world’s Mohan, M. and Kandya, A. (2007) ‘An analysis of the annual and poorest cities. Available at: http://www.who.int/en/news-room/ seasonal trends of air quality index of Delhi’, Environmental detail/12-05-2016-air-pollution-levels-rising-in-many-of-the- Monitoring and Assessment, 131(1–3), pp. 267–277. doi: 10.1007/ world-s-poorest-cities (Accessed: 27 June 2018). s10661-006-9474-4. WHO (2016b) Health risk assessment of air pollution: General Okello, N. and Allan, C. (2015) ‘The Richards Bay Clean Air principles. Available at: http://www.euro.who.int/en/health- Association : A Case Study for Success in Participatory Air Quality topics/environment-and-health/air-quality/publications/2016/ Management’, Clean Air Journal, 25(1), pp. 34–39. Available at: health-risk-assessment-of-air-pollution.-general- http://dx.doi.org/10.17159/2410-972X/2015/v25n1a2. principles-2016.

Pohlert, T. (2018) ‘Non-Parametric Trend Tests and Change-Point WHO (2016c) ‘WHO | Air pollution’, World Health Organization. Detection’, R package, p. 26. doi: 10.13140/RG.2.1.2633.4243. Available at: http://www.who.int/topics/air_pollution/en/.

Rai, P. K. (2015) ‘Multifaceted health impacts of Particulate World Health Organization (2016) ‘Ambient Air Pollution: A Matter ( PM ) and its management : An overview’, Environmental global assessment of exposure and burden of disease’, World Skeptics and Critics, 4(1), pp. 1–26. Health Organization, pp. 1–131. doi: 9789241511353.

RBCAA (2018) 2017 Richards Bay Clean Air Association (RBCAA) Annual Air Quality Report. Richards Bay.

CLEAN AIR JOURNAL 89 Volume 28, No 2, 2018